1. 1 Programmes of Measures and Standards Freshwater Morphology Workshop Pilot Study - Risk Assessment Refinement Lorraine Houston Sheila Downes 4 th April 2008

2. 2 Risk Assessment Refinement Use field data to refine Article 5 risk assessment pressure thresholds Aerial Imagery Capture and Processing Derive Channel TypologiesDevelop Remote Sensing Methods Fieldwork Contracts 2006 and 2007 Morphology and Biology Field Data Decision Support Tool Programmes of Measures Populate a morphology database Developing the Morphological Assessment Process RISK ASSESSMENT CLASSIFICATION & MONITORING PoMS STUDY

3. 3 Refinement of Risk Assessment Thresholds Most Significant Pressures Channelisation Intensive Land Use These were also the pressures that had most uncertainty in Article 5 Risk Assessment Investigation undertaken to improve our confidence in assessing the potential risk of these pressures How much morphological and biological impact do these pressures actually have?

4. 4 PoMS Pilot Study- Site Selection and Fieldwork

5. 5 Fieldwork A full programme of investigative fieldwork was commissioned during 2007 on a range of pilot waterbodies throughout Ireland. This fieldwork was aimed at acquiring enough morphological and biological data To refine the pressure thresholds Facilitate work required in other parts of the PoMS Study Provide Field based verification data for the GIS tool development work

6. 6 Site Selection The priority in site selection was pilot waterbodies where channelisation and intensive land use pressures were identified as the only pressure, morphological or otherwise posing risk of failure to meet “Good Ecological Status” objectives by 2015. Why? This would allow attribution of observed impact to these pressures since no other pressures are acting on the waterbodies. In contrast, sites which are deemed to be of “High Status” were also selected so that a range of pressure thresholds could be observed in the field.

7. 7 Pilot Waterbodies – Site Selection Criteria Sites selected were spread across Ireland Sites Categorised A-C had the most significant pressures with only intensive land use or channelisation putting them “At Risk” Sites D – F were selected as “High Status” sites Sites G, H, J (observed morph impact by EPA) & EPA Surveillance sites EHS Sites

8. 8 Fieldwork Methodologies – Morphology Rapid Assessment Technique (R.A.T) (Based on Observed Impact) 500m stretch WFD Classes related to R.A.T score: > 0.8 = High 0.6 – 0.8 = Good 0.4 – 0.6 = Moderate 0.2 – 0.4 = Poor <0.2 = Bad Morphological Impact Assessment System (MImAS) (Based on recording the pressures that could cause impact) 500m stretch Scores are based on the amount of capacity a river has to accept morphological change. If more than 15% capacity is used up, river is deemed to be at risk of less than good status.

9. 9 Fieldwork Methodologies – Biology Biology surveys were undertaken by Shannon IRBD Project Staff and Aquatic Services Unit, UCC. Two types of biological surveys were undertaken:  Biological Q Assessment (Macroinvertebrates)  CBAS (Canonical Correspondence Analysis Based Assessment System) (Macrophytes) Biological Q Assessment The assessment carried out was a modified version of the EPA technique. No Physico-chemical analysis was carried out at the sites. Macrophytes and Filamentous Algae were recorded at the sites mainly on a presence or absence basis. Fieldwork focused on the two-minute Macroinvertebrate kick sampling technique together with two minute stone wash and pond net survey. All results were recorded in the standard EPA Rivers Ecological Assessment Field Sheet.

10. 10 Q Rating Comparison with RAT scores The existing Q system has been modified by the EPA to the equivalent WFD status class Sites with a Q score less than Q4 are deemed “less than Good” in terms of the macroinvertebrate component of WFD ecological status.

11. 11 Q Rating Comparison with R.A.T scores R.A.T Scores and Corresponding Q Scores – All Pilot Waterbodies (Ireland) In general, there is no conflict between Q score and R.A.T score Sites with R.A.T scores less than 0.6 (less than good status) generally have Q scores less than 4 (less than good status in terms of macroinvertebrates) Good or HighLess than Good Sites with lower Q Scores, whilst also have low R.A.T Scores due to modifications or recent dredging, are a result of a combination of pollution pressures, not just morphology (Heavily Modified)

12. 12 Fieldwork Methodologies - Biology CBAS The river CBAS survey method was developed through the North South Share Project as a tool to assign ecological status based on presence or absence of in channel macrophyte species A score is calculated based on impact metrics associated with nutrient loading (Soluble Reactive Phosphate, Nitrates and Ammonia) and hydromorphology (Substrate, Dissolved oxygen and pH). A high impact metric indicates a deviation from reference condition. The impact metrics are generated based on the type of macrophytes that are present in the river as an indicator of the nutrient and morphological condition.

13. 13 The combination of these metrics generate an overall CBAS score which is related to WFD Status classes. Two overrides exist with the tool; 1. >50% Alien Species Automatically reduces status to “bad” 2. >50% Eutrophic species Site is automatically deferred to at most “Poor Status” CBAS

14. 14 CBAS Through consultation with Ian Dodkins (one of the tool developers) it was established that the SUBSTRATE impact metric would be a good indicator of morphological alteration. A high deviation from reference condition for the SUBSTRATE metric indicates a higher level of siltation due to morphological impact in the channel.

15. 15 Therefore, for the purposes of this study in terms of morphology it was decided that the SUBSTRATE impact metric would be looked at in relation to the R.A.T score. Study Findings 1.Sites with high Substrate Impact metrics have low R.A.T scores. The majority of sites with Substrate Impact Metrics greater than 5 have R.A.T scores less than 0.6 (Less than good morphology status) 2.A subset of the sites were subject to intensive land use pressures in isolation. The R.A.T score for these sites drop below 0.6 when the substrate impact metric is greater than 8 3.Sites subject to channelisation pressures in isolation have R.A.T scores less than 0.6 when the Substrate impact metric is greater than 4. Good or HighLess than Good Increasing Impact

16. 16 Biology Findings The observed relationship between biological data and R.A.T scores has indicated that morphological pressure, can contribute to overall impact on biology and therefore ecological status. In general, sites with R.A.T scores less than 0.6 also have less than good Q scores, albeit a combination of pressures are acting on sites with very low Q scores Similarly, high levels of siltation affecting macrophyte populations are reflected by less than good R.A.T scores. Whilst this is a result of a combination of pressures, the associated sustainable level of channelisation and ILU must now be identified to refine the Article 5 Risk Assessment.

17. 17 Survey Sites Optimising Risk Assessment Thresholds Good or High StatusLess than Good Status

18. 18 Survey Sites At Risk? YES Low R.A.T Score? YES At Risk? NO Good R.A.T Score? YES Optimising Risk Assessment Thresholds

19. 19 Survey Sites At Risk? No Low R.A.T Score? YES Optimising Risk Assessment Thresholds

20. 20 Survey Sites At Risk? Yes Good R.A.T Score? YES But conservative Optimising Risk Assessment Thresholds

21. 21 Optimising Risk Assessment Thresholds R.A.T Score =0.6 Risk Assessment Threshold Not at Risk in R.A At Risk in R.A At Risk in (R.A.T) Field Survey Not at Risk in (R.A.T) Field Survey Quadrant 1 Sites classified: At risk in Risk Assessment + At risk in R.A.T Survey = Risk Assessment is Correct Quadrant 2 Sites classified: At risk in Risk Assessment + Not At risk in R.A.T Survey = Risk Assessment is Conservative Quadrant 4 Sites classified: Not At risk in Risk Assessment + Not At risk in R.A.T Survey = Risk Assessment is Correct Quadrant 3 Sites classified: Not At risk in Risk Assessment + At risk in R.A.T Survey = Risk Assessment is Incorrect

22. 22 Channelisation Article 5 Risk Assessment Threshold between ‘at risk’ and ‘not at risk’ = 15% All rivers with >15% channelisation were identified as “probably at risk” of failing to meet Good Status by 2015 Risk assessment was capped at “probably at risk” due to uncertainties as to the long term impact of dredging activities Uncertainties Investigated in the PoMS Study: 1.Is 15% a good reflection of how channelisation can impact the status of a waterbody? 2.Does river response vary according to channel type? 3.Does watercourse maintenance of drained rivers impact ability to recover morphologically and as a subsequence ecologically?

23. 23 Is 15% a good reflection of how channelisation can impact the status of a waterbody? 15% Too Conservative Good or HighLess than Good A general trend is evident Increasing R.A.T score with Decreasing % Channelisation in waterbody

24. 24 Is 15% a good reflection of how channelisation can impact the status of a waterbody? 50% 71% of sites are classified correctly using the 50% threshold for channelisation Several sites with 10% channelisation or less have R.A.T scores greater than 0.8 (high status) It is recommended that the 15% threshold between good and less than good status is increased to 50%. Good or HighLess than Good A general trend is evident Increasing R.A.T score with Decreasing % Channelisation in waterbody

25. 25 However, there are still cases where waterbodies with high percentage channelisation have high R.A.T scores - conservative There are also cases where sites with low percentage channelisation have less than good R.A.T scores - incorrect This raises the questions: Does channel type affect morphological response to channelisation? Does watercourse maintenance affect morphological recovery? Is 15% a good reflection of how channelisation can impact the status of a waterbody?

26. 26 BedrockStep Pool Cascade Pool Riffle Lowland Meandering R.A.T Types Channel Type Diagram: Rosgen

27. 27 Does Channel Response to Channelisation Vary According to Channel Type? Good or HighLess than Good Upland rivers are less sensitive to channelisation pressures Tend to have high R.A.T scores (greater than 0.8) % Channelisation V R.A.T Score – According to Channel Type

28. 28 Does Channel Response to Channelisation Vary According to Channel Type? Good or HighLess than Good It could be argued that a higher threshold between good status and less than good status could be applied to upland rivers However it is recommended that 50% is applied throughout since the majority of rivers subjected to drainage are lowland meandering and pool riffle. The majority of rivers with high percentages of channelisation and R.A.T scores below 0.6 are lowland meandering and pool riffle. Lowland rivers are more sensitive to channelisation

29. 29 Does watercourse maintenance affect morphological recovery? Percentage of Maintained Rivers with Good or Less than Good R.A.T Scores Percentage of Non Maintained Rivers with Good or Less than Good R.A.T Scores Research indicates that continual watercourse maintenance in drained rivers impedes the recovery process Maintenance records for the survey sites subject to channelisation were sought from the Office of Public Works (OPW) in Ireland and DARD Rivers Agency in NI.

30. 30 Does watercourse maintenance affect morphological recovery? Maintained RiversNon Maintained Rivers Measures to mitigate against this impact should be included in the Programmes of Measures within River Basin Management Plans. Whether a channel is maintained or not should be accounted for in the risk assessment. A higher (less stringent) threshold should be applied to rivers that are not maintained. However, maintenance records are not readily available at present for all rivers. This dataset should be improved with a view to refining the risk assessment further in the second RBMP cycle.

31. 31 15% (Old) Channelisation Recommendations EXAMPLE – SHANNON IRBD AT RISK NOT AT RISK

32. 32 15% (Old) Channelisation Recommendations – Increase Threshold 50% Proposed

33. 33 Channelisation Recommendations OPW Drained Channels 15% 50%

34. 34 Channelisation Recommendations The risk assessment threshold between ‘at risk’ and ‘not at risk’ should be increased from 15% to 50% - Discussion It could be argued that a higher threshold between good status and less than good status could be applied to upland rivers. However it is recommended that 50% is applied throughout since the majority of rivers subjected to drainage are lowland meandering and pool riffle. Measures to mitigate against watercourse maintenance should be included in the Programmes of Measures within River Basin Management Plans. Whether a channel is maintained or not should be accounted for in the risk assessment. A higher (less stringent) threshold should be applied to rivers that are not maintained. However, maintenance records are not readily available at present for all rivers. This dataset should be improved with a view to refining the risk assessment further in the second RBMP cycle.

35. 35 Intensive Land Use Article 5 Risk Assessment Threshold between ‘at risk’ and ‘not at risk’ = 30% All rivers with 30 – 70% intensive land use were identified as “probably at risk” of failing to meet Good Status by 2015 All rivers with > 70% intensive land use were identified as “at risk” of failing to meet Good Status by 2015 The % ILU was calculated on GIS as the length of river (within 50m of the river banks) flanked by ILU zones as a proportion of the total river length The ILU zones included: Forestry Arable Land Urban Fabric Exploited Peat Land As depicted by the Corine 2000 GIS Dataset.

36. 36 Uncertainties Investigated in the PoMS Study: 1.Is 30% a good reflection of how ILU can impact the status of a waterbody? 2.Should Improved Grassland be included as an Intensive Land Use type? 3.Does river response vary according to channel type? 4.Should calculation of percentage ILU be based on the upstream catchment or along the river itself? Intensive Land Use

37. 37 Is 30% a good reflection of how ILU can impact the status of a waterbody? Should Improved Grassland be included as a Land Use Type in addition to forestry, arable, peat and urban? Would this improve the relationship? Intensive Land Use is an Indirect Pressure Relationship between % ILU in a waterbody and R.A.T score (observed impact) is not easily defined Most sites had good or high R.A.T scores regardless of % ILU

38. 38 Should Improved Grassland be included as a Land Use Type in addition to forestry, arable, peat and urban? There is an improved relationship It is recommended that Improved Grassland is included in the risk assessment However, relationship is not strong, threshold must be optimised Can impact river morphology at a local scale in the form of cattle poaching and removal of riparian zones. It can also impact more indirectly e.g. overgrazing which increases soil run-off to rivers and increases sediment movement within the system.

39. 39 Does River Response Vary According to Channel Type? Pool Riffle and Lowland Meandering most sensitive to ILU pressures 30% threshold too conservative

40. 40 Should calculation of percentage ILU be based on the upstream catchment or along the river itself?

41. 41 Should calculation of percentage ILU be based on the upstream catchment or along the river itself? Relationship not significantly different using area method Recommend that linear method is used for risk assessment Revisit for 2 nd RBMP cycle

42. 42 Intensive Land Use - Optimising the Risk Assessment Threshold Findings / Recommendations So Far: Improved Grassland should be included in the risk assessment as a land use type Lowland rivers are more sensitive to ILU pressures Calculation of % ILU is based on river length Relationship between % ILU and R.A.T is still unclear Must select a risk assessment threshold that: Maximises the no. waterbodies assigned risk correctly Minimises the no. waterbodies assigned risk incorrectly Maintains a conservative approach but provides a truer reflection of morphological impact on the ground than Article 5 risk assessment

43. 43 Intensive Land Use - Optimising the Risk Assessment Threshold R.A.T Score =0.6 % ILU Risk Assessment Threshold Not at Risk in R.A At Risk in R.A At Risk in (R.A.T) Field Survey Not at Risk in (R.A.T) Field Survey Quadrant 1 Sites classified: At risk in Risk Assessment + At risk in R.A.T Survey = Risk Assessment is Correct Quadrant 2 Sites classified: At risk in Risk Assessment + Not At risk in R.A.T Survey = Risk Assessment is Conservative Quadrant 4 Sites classified: Not At risk in Risk Assessment + Not At risk in R.A.T Survey = Risk Assessment is Correct Quadrant 3 Sites classified: Not At risk in Risk Assessment + At risk in R.A.T Survey = Risk Assessment is Incorrect

44. 44 Does River Response Vary According to Channel Type? 30% threshold too conservative

45. 45 Does River Response Vary According to Channel Type? 60-70% threshold?

46. 46 Intensive Land Use - Optimising the Risk Assessment Threshold No. Sites Assigned Risk Correctly – too low Too Conservative No. Sites Assigned Risk Correctly – high But Incorrect high and Not Conservative enough

47. 47 Intensive Land Use - Optimising the Risk Assessment Threshold

48. 48 30% (Old) Intensive Land Use Recommendations EXAMPLE – SHANNON IRBD

49. 49 30% (Old) Intensive Land Use Recommendations EXAMPLE – SHANNON IRBD AT RISK NOT AT RISK 70% Proposed

50. 50 Intensive Land Use Recommendations The risk assessment threshold between ‘at risk’ and ‘not at risk’ should be increased from 30% - Discussion Improved Grassland should be included as an ILU Type Calculation of % ILU should be based on the length of river flanked by ILU zones as per the Article 5 Risk Assessment methodology. However, this should be reviewed for the second RBMP cycle following further research on the spatial impact of ILU pressures. This should pick up more direct ILU pressures acting on the river itself. Catchment wide pressures such as overgrazing can still be accounted for using expert judgement input in Programmes of Measures. Whilst lowland meandering and pool riffle rivers are found to be more sensitive to ILU pressures, it is recommended that all river types should be included in the risk assessment for the first RBMP and reviewed for the second RBMP cycle when GIS based channel typology tool is completed.

51. 51 Q & A  Channelisation Do you agree with the increase in pressure threshold from 15% to 50% Feedback on other findings – watercourse maintenance, channel type?  Intensive Land Use Do you agree with increase in pressure threshold from 30% to 70% Feedback on other findings – improved grassland inclusion, channel type, method?